The primary objective of this proposal is to extend our earlier investigations into the perception of phonetic features and segments in the very young, prearticulate infant. Information derived from studies of this nature have implications not only for biological foundations of language but also for current theories of speech perception. To accomplish this objective, specific studies have been proposed to investigate perception of some of the phonetic contrasts found in most human languages. A feature of these studies is that they will also provide information regarding the manner in which these phonetic contrasts are perceived, that is, in a linguistic or auditory mode. Studies are also planned to investigate the role of early receptive experience on infant speech perception, and the ability of infants to organize features. Additional goals are to continue our investigations of the properties of auditory and phonetic feature detectors that are assumed to underlie the perception of speech, to investigate the structure of the speech stimulus and its relation to individual processing strategies, and finally to investigate the availability and durability of the auditory and phonetic information in the speech stimulus. The infant studies will use a procedure that makes presentation of the stimulus contingent upon high-amplitude sucking. When the infant's sucking rate indicates a decrement in the response-producing properties of the initial stimulus, a second stimulus is introduced. Change in the sucking rate (given appropriate controls) is taken as evidence for discrimination. With appropriate manipulations of the synthetic speech stimuli, information related to the nature of processing and the role of experience can be forthcoming. A selective adaptation procedure will be used to investigate the properties of feature detectors. Studies related to the structure of the speech stimulus and the availability of auditory and phonetic information will use the recently developed paradigm of Garner and the stimulus matching paradigm of Posner, respectively. These latter studies will use adult listeners.